Glenn Roisman

Who they are

The first was in 2002, on "earned secure" attachment. The category describes an adult who narrates a difficult childhood coherently, and the popular reading is that such a person had a hard start and overcame it. Roisman and colleagues had access to a group followed since infancy, so they could check. The adults who now told a hard story coherently had not, on the record, had harder childhoods. Observed at the time, they had received some of the most supportive maternal parenting in the sample. Their recommendation was to reserve the term for adversity established prospectively, from observation rather than recall.

The second, in 2007, is the more consequential. Attachment research has two adult instruments. The Adult Attachment Interview takes about an hour and scores how coherently a person can narrate their childhood. Self-report questionnaires ask how a person feels about closeness now. Both call the good outcome secure, and for two decades the field proceeded as though they were two windows onto one thing.

Roisman and colleagues pooled every study that had run both on the same people, nearly a thousand adults, and then ran a fresh study with the best available version of each. The overlap was, in their own words, trivial to small. Under one part in a hundred. The only thread above negligible was a weak link between self-reported avoidance and a dismissing interview style.

That result does not show either instrument is worthless. It shows they are not measuring the same construct while sharing a vocabulary, which is why an article citing Bowlby at the top and administering a quiz at the bottom is joining two literatures that do not join.

He is also a co-author of the 2023 preregistered reanalysis of two landmark cohorts, which found infant attachment predicting academic outcomes about as well as socioemotional ones.

Key papers

Key papers

Roisman, G. I., Padrón, E., Sroufe, L. A., & Egeland, B. (2002). Child Development, 73(4), 1204-1219. First author. The earned-secure study, run inside a cohort followed from infancy, so a recalled childhood could be set against a recorded one. Adults who now narrated a difficult childhood coherently had not, on the prospective record, had harder childhoods, and had been observed receiving some of the most supportive maternal care in the sample. The authors recommend reserving the term for adversity established prospectively.

Roisman, G. I., Holland, A., Fortuna, K., Fraley, R. C., Clausell, E., & Clarke, A. (2007). Journal of Personality and Social Psychology, 92(4), 678-697. First author. The convergence study. Every available dataset in which the same adults had taken both the Adult Attachment Interview and a self-report attachment questionnaire was pooled, close to a thousand people, and a fresh study was run with the best version of each instrument. The overlap was, in the authors' own words, trivial to small. The only association above negligible ran between self-reported avoidance and a dismissing interview style.

Caldo, P. D., and colleagues (2023). Development and Psychopathology, 37(1), 147-163. Co-author. A preregistered test of the Strange Situation's predictive validity across two landmark cohorts, in which infant attachment predicted academic outcomes about as well as it predicted socioemotional ones.

Key ideas

Check the construct before building on it. Two measures sharing a word is not evidence they share a referent.

Prospective beats retrospective. The earned-secure result is what happens when a recalled childhood meets a recorded one.

Preregister the test of your own field's flagship measure, which is what 2023 did.

Concepts on this site

Construct Validity

Earned Secure Attachment

---